Exploring Undergraduate AI Perceptions: Knowledge, Enthusiasm, and Concerns

As Artificial Intelligence (AI) develops and grows its presence in society, college students are increasingly interacting with AI and utilizing tools like ChatGPT as part of their education. Particularly in STEM fields, educators themselves are incorporating AI by encouraging its use as an assistive tool for coursework or designing courses that teach about its inner workings. Understanding students' perceptions and knowledge of AI can help educators know whether students will embrace learning in AI-heavy environments, as well as which student concerns they should acknowledge. Our study uses both quantitative and qualitative data from undergraduate CMNS (College of Computer, Mathematical, and Natural Sciences) students at the University of Maryland, College Park to explore students' perceived knowledge, enthusiasm, and concerns over AI. Our data was collected via a survey administered via email to undergraduates and subsequent focus group interviews with these students about their relationship with AI. Survey findings indicated that students were confident in their knowledge of AI and related competencies, as well as enthusiastic about learning and using AI. Students also highly believed in the need for standards and testing for AI systems to curtail risks. There was a positive correlation between perceived knowledge and enthusiasm of AI, but no correlation between knowledge and concerns. In interviews, students' main uses of AI were summarizing information, creating practice problems, and writing assistance. Popular concerns included academic dishonesty, overreliance on AI tools, and fabricated information in outputs.